[Bug build/18796] New: build fails for --disable-mathvec

[Bug build/18796] New: build fails for --disable-mathvec

[swb@michael-joost.de]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

            Bug ID: 18796
           Summary: build fails for --disable-mathvec
           Product: glibc
           Version: 2.22
            Status: NEW
          Severity: normal
          Priority: P2
         Component: build
          Assignee: unassigned at sourceware dot org
          Reporter: swb@michael-joost.de
                CC: carlos at redhat dot com
  Target Milestone: ---

When compiling for a non-SSE x86 CPU, hence --disable-mathvec, that mvec lib is
not compiled, but the generated soversions.mk contains an entry:

  libmvec.so-version=.1
  all-sonames+=libmvec=libmvec.so$(libmvec.so-version)

which is later (make install) consulted by scripts/testinstallation.pl, causing
this error:

.../i586-pc-linux-gnu/bin/ld: cannot find -lmvec
collect2: error: ld returned 1 exit status
Execution of i586-pc-linux-gnu-gcc failed!

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[jsm28 at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

Joseph Myers <jsm28 at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andrew.n.senkevich at gmail dot co
                   |                            |m

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[andrew.n.senkevich at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

Andrew Senkevich <andrew.n.senkevich at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |joseph at codesourcery dot com

--- Comment #1 from Andrew Senkevich <andrew.n.senkevich at gmail dot com> ---
Joseph,

how I see it is needed to fix test-installation.pl itself with excluding -lmvec
if according case found with help of grep on config.make. 
Is it ok?

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[andrew.n.senkevich at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

Andrew Senkevich <andrew.n.senkevich at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[joseph at codesourcery dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #2 from joseph at codesourcery dot com <joseph at codesourcery dot com> ---
Please send your patch to libc-alpha as usual with the rationale for the 
implementation approach you choose.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[swb@michael-joost.de]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #3 from Michael Joost <swb@michael-joost.de> ---
Just a thought: have you considered to adapt the generation of the
soversions.mk to drop that entry, instead of modifying the test-installation
script ?

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[andrew.n.senkevich at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #4 from Andrew Senkevich <andrew.n.senkevich at gmail dot com> ---
(In reply to Michael Joost from comment #3)
> Just a thought: have you considered to adapt the generation of the
> soversions.mk to drop that entry, instead of modifying the test-installation
> script ?

That entry needed in soversions.i for further generation of definition
MODULE_libmvec in ./libc-modules.h needed for IS_IN (libmvec) proper unfolding.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[joseph at codesourcery dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #5 from joseph at codesourcery dot com <joseph at codesourcery dot com> ---
On Thu, 13 Aug 2015, andrew.n.senkevich at gmail dot com wrote:

> https://sourceware.org/bugzilla/show_bug.cgi?id=18796
> 
> --- Comment #4 from Andrew Senkevich <andrew.n.senkevich at gmail dot com> ---
> (In reply to Michael Joost from comment #3)
> > Just a thought: have you considered to adapt the generation of the
> > soversions.mk to drop that entry, instead of modifying the test-installation
> > script ?
> 
> That entry needed in soversions.i for further generation of definition
> MODULE_libmvec in ./libc-modules.h needed for IS_IN (libmvec) proper unfolding.

In addition, I think gnu/lib-names.h should only very (among builds for 
the same architecture / OS) with the ABI variants listed in 
$(abi-variants), so that the same set of headers can be shared between 
different glibc builds, not depending on configure options such as whether 
libmvec is enabled or not.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #6 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU C Library master sources".

The branch, master has been updated
       via  a08e80d1143f6b0386d5bc8cc7b8ed576091dbf3 (commit)
      from  9b9d20faf6185f5dcb36cde48ca3c9db866086ee (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=a08e80d1143f6b0386d5bc8cc7b8ed576091dbf3

commit a08e80d1143f6b0386d5bc8cc7b8ed576091dbf3
Author: Andrew Senkevich <andrew.senkevich@intel.com>
Date:   Wed Aug 19 16:54:28 2015 +0300

        [BZ #18796]
        * scripts/test-installation.pl: Don't add -lmvec to build options if
        libmvec wasn't built.

-----------------------------------------------------------------------

Summary of changes:
 ChangeLog                    |    6 ++++++
 scripts/test-installation.pl |   13 ++++++++++++-
 2 files changed, 18 insertions(+), 1 deletions(-)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[joseph at codesourcery dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #7 from joseph at codesourcery dot com <joseph at codesourcery dot com> ---
Remember if checking in a fix for a bug to add the bug to the list of 
fixed bugs in NEWS and close the bug as fixed as well.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #8 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU C Library master sources".

The branch, release/2.22/master has been updated
       via  9031106ea063f0476bdabf3f5ec22758cdcf987b (commit)
      from  a34d1c6afc86521d6ad17662a3b5362d8481514c (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=9031106ea063f0476bdabf3f5ec22758cdcf987b

commit 9031106ea063f0476bdabf3f5ec22758cdcf987b
Author: Andrew Senkevich <andrew.senkevich@intel.com>
Date:   Wed Aug 19 19:02:32 2015 +0300

        [BZ #18796]
        * scripts/test-installation.pl: Don't add -lmvec to build options
        if libmvec wasn't built.
        * NEWS: Mention this fix.

-----------------------------------------------------------------------

Summary of changes:
 ChangeLog                    |    7 +++++++
 NEWS                         |    2 +-
 scripts/test-installation.pl |   13 ++++++++++++-
 3 files changed, 20 insertions(+), 2 deletions(-)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[andrew.n.senkevich at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

Andrew Senkevich <andrew.n.senkevich at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #9 from Andrew Senkevich <andrew.n.senkevich at gmail dot com> ---
(In reply to joseph@codesourcery.com from comment #7)
> Remember if checking in a fix for a bug to add the bug to the list of 
> fixed bugs in NEWS and close the bug as fixed as well.

Is it ok to commit such NEWS fix with no ChangeLog entry?

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[joseph at codesourcery dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #10 from joseph at codesourcery dot com <joseph at codesourcery dot com> ---
On Wed, 19 Aug 2015, andrew.n.senkevich at gmail dot com wrote:

> > Remember if checking in a fix for a bug to add the bug to the list of 
> > fixed bugs in NEWS and close the bug as fixed as well.
> 
> Is it ok to commit such NEWS fix with no ChangeLog entry?

Yes, there is no need for ChangeLog entries for fixes to NEWS or to 
ChangeLog files.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

--- Comment #11 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU C Library master sources".

The branch, master has been updated
       via  1f3be1988bce6756ad8cd37386eedef95aa0f2ed (commit)
      from  48497aba8ec7887dd505e37aa95ab52348fa8cb2 (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=1f3be1988bce6756ad8cd37386eedef95aa0f2ed

commit 1f3be1988bce6756ad8cd37386eedef95aa0f2ed
Author: Andrew Senkevich <andrew.senkevich@intel.com>
Date:   Wed Aug 19 19:48:52 2015 +0300

    Mention BZ #18796 fix in NEWS.

-----------------------------------------------------------------------

Summary of changes:
 NEWS |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug build/18796] build fails for --disable-mathvec

[jsm28 at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18796

Joseph Myers <jsm28 at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |2.23

-- 
You are receiving this mail because:
You are on the CC list for the bug.